Phospholipids (PLs) are unique lipids that play a crucial role in food systems due to their amphiphilic nature, contributing to emulsion formation, structural stability, and texture modulation in a wide range of products (1–3). Dietary PLs play a particular role as essential components of cellular membranes and lipid metabolism (4), as well as other emerging health-promoting functionalities. They contribute to the intake of choline and essential fatty acids and have been associated with beneficial effects on brain, liver, and cardiovascular health (5).

In fact, increasing evidence indicates that dietary PLs support brain health and cognitive function, as they are integral components of neuronal membranes and contribute to synaptic plasticity, neurotransmission, and age-related cognitive maintenance (6–8). Specific PLs classes, particularly phosphatidylserine (PS), phosphatidylcholine (PC), and plasmalogens (PL), and marine-derived PLs-bound ω-3 fatty acids, have been associated with neuroprotective effects, including reported improvements in memory, learning capacity, and executive function (9–11).

Moreover, PLs and their metabolites, particularly lysophospholipids, were recognized as key modulators of inflammatory and immune system response (12). They act as bioactive lipid mediators that influence immune homeostasis and the inflammatory process with implications for chronic inflammatory conditions and immune-related disorders (12–15). Specific PLs classes, such as Plasmalogens and phosphatidylserine, further contribute to antioxidant and anti-inflammatory protection by enhancing cellular resilience against oxidative stress, as extensively discussed in recent reviews (10, 16–18). From a technological perspective, PLs are widely used in food and nutraceutical formulations as natural emulsifiers, stabilizers, and delivery systems for lipophilic bioactives, due to their technological roles (1, 19, 20). These various physiological and technological functions are directly related to the dietary sources of PLs, which vary significantly in their composition and structural features.

The main food sources of PLs include eggs, dairy products, soybeans, fish, and oilseeds, which represent rich dietary sources of these compounds (5). This source diversity directly influences the functional properties of the PLs and their extraction methods (1, 20). This diversity is primarily controlled by three main factors, which are: fatty acid composition (length and saturation level), head group structure, which determines its charge and hydrophobicity (choline, serine, inositol …), and extraction approaches (21, 22). The variation in these factors affects membrane fluidity, emulsifying capacity, and oxidative stability (23). For instance, the ability of PLs as an emulsifier, which allows them to coat the oil droplets, is highly correlated with different factors such as the origin (source) of the PLs, the proportion of PC to other PLs, as well as the saturation of fatty acids (1). PLs from peony seeds, rich in PC (78.49%) and lysophosphatidylcholine (LPC) (8.11%), along with a considerable proportion of polyunsaturated fatty acids (PUFA: 88.84%), have an effective emulsification ability and produce uniform oil droplets (20). Meanwhile, soybean PLs, the most common commercial source of PLs with high content of PC, phosphatidylethanolamine (PE), phosphatidylinositol (PI), phosphatidic acid (PA), and high PUFA, have a strong ability to stabilize the emulsified system, mainly through protein complexes (23). They are tightly associated with neutral lipids, proteins, and carbohydrates, making degumming followed by solvent partitioning or supercritical CO₂ de-oiling necessary to disrupt the lipid-protein interactions and selectively enrich PLs (24, 25). In contrast, egg yolk PLs, also frequently analyzed PLs type embedded within complex lipoprotein assemblies and membranes, contain a high percentage of PC (~72%), followed by PE, PI, PL, and sphingomyelin (SM) (26). There is also a diverse fatty acid profile that includes long-chain PUFA such as docosahexaenoic acid (DHA) and eicosapentaenoic (EPA) (10). The strong association of PLs with proteins and cholesterol in egg yolk necessitates solvent systems or emerging enzymatic and membrane-based approaches to enhance extraction efficiency while preserving PLs integrity (27, 28). Meanwhile, marine PLs are also valuable sources with a distinctive fatty acid composition that is particularly rich in omega-3 long-chain PUFA such as EPA and DHA. Methods to process and extract PLs from marine sources are more specialized and require milder conditions to safeguard their oxidative stability and biological activity (29–31). The significant source-dependent structural and compositional diversity directly complicates the extraction, separation, and analytical characterization of PLs across diverse dietary matrices.

One of the key challenges in PLs research is the lack of standardization of analysis procedures across different food matrices, which limits progress compared to neutral lipid studies. Besides their limited amounts and source-dependent variability, PLs analysis efficiency and reliability are additionally challenged by food matrix complexity and processing methods, which directly impact their extraction as well as their characterization (13, 32, 33). Components such as proteins, neutral lipids, polysaccharides, and pigments can strongly interact with PLs, lowering extraction efficiency and increasing the co-extracting interfering compounds, as seen in protein-rich foods like dairy and egg yolk (26, 34) or pigment-rich plant materials (35). Consequently, extraction strategies and cleanup methods often need to be adapted to the specific food matrix and the intended application to obtain reliable PLs data (36–38). These issues are further impaired by the lack of standardized analytical procedures across studies.

PLs extraction methods range from conventional solvent-based procedures (39, 40) to supercritical CO₂, enzyme-assisted, or ultrasound-assisted extractions (17, 18, 30, 31, 41–43). These methodologies limit the comparability of results across studies. Moreover, chromatographic characterization of PLs also lacks consistency with methods ranging from HPLC–MS to RP-HPLC–MS/MS, in addition to the limited consensus with key isomer distinctions (30, 42, 44–46). Furthermore, other challenges arise in quantifying PLs due to the broad reference standards and calibration used, which add a further layer of analytical inconsistencies in the current literature. However, updated methodologies have emerged that can address these traceability concerns, including HILIC-ESI-AIF-MS coupled with ³¹P NMR or ICP-MS (5, 44, 47–49). Yet, in a recent study on milk PLs, it has been reported that regardless of the use of advanced LC–MS systems for detection, differences in extraction and MS conditions are still altering the result substantially (50). These analytical variations restrict the reproducibility of research findings and hinder broader scientific and industry development.

There is a lack of consensus on the extraction, separation, and detection methodologies, which leads to results that are generally non-comparable with low quality, negatively affecting database creation, new product development, and process optimization. Our review comprehensively provides a synthesis and critical evaluation of existing extraction, separation, purification, identification, and quantification methods for PLs from different foods as well as other biological matrices.

Structure-wise, PLs are amphiphilic molecules, consisting of two hydrophobic fatty acid tails and a phosphate-based polar head group (51). Accordingly, PLs can be divided into subclasses based on their headgroups (Figure 1). The primary food glycerophospholipids are phosphatidylcholine, phosphatidylethanolamine, phosphatidylinositol, phosphatidylserine, phosphatidylglycerol (PG), and phosphatidic acid. All glycerophospholipids share a phosphatidic acid backbone, which includes a glycerol backbone with two fatty acyl esters and a phosphate group, with a phosphodiester-linked headgroup to choline, ethanolamine, inositol, serine, or glycerol (Figure 1) (11, 27).

PC is zwitterionic (bearing both positive and negative charges yet net neutral at physiological pH) and is generally the most abundant in egg yolk and muscle tissues (Figure 2) (41). Moreover, PE is also zwitterionic and may match PC. On the other hand, PI and PS are anionic at physiological pH and are often less abundant, although PS is found at ~2–11% of bovine milk PLs (27). PG is also anionic and rich in plants and microbes; its dimer, cardiolipin (CL), is abundant in heart tissue but found only in trace amounts in foods like beef (44, 52). Sphingophospholipids, predominantly sphingomyelin (SM), are also found in foods (Figure 2). SM has a ceramide backbone that includes sphingosine with a fatty acid and a phosphocholine headgroup (mirroring PC) and is highly abundant in milk and egg (27). Lysophospholipids-LPC and LPE are mono-acyl derivatives of PC and PE that can be formed during digestion or food processing. Though they generally existed at lower concentrations, these Lysophospholipids still need to be accounted for.

Due to the amphiphilic nature, PLs usually assemble into membranes, micelles, and liposome-like structures and exhibit strong interactions with proteins, carbohydrates, and other food matrix components (27). These interactions create challenging conditions for their release from food matrices compared to neutral lipids (53–55). Moreover, most glycerophospholipids share a common phosphatidyl backbone and differ only slightly in head-group chemistry or fatty-acyl composition, resulting in extensive isobaric and isomeric overlap (56). This high degree of structural similarity among PLs classes represents a fundamental obstacle to their selective isolation and accurate characterization (Figure 2) (3). The highly variable fatty acyl composition also contributes to this complexity; for example, marine PLs are rich in omega-3 species while dairy and meat PLs are dominated by saturated and monounsaturated species (10, 54, 57). Additional variations in acyl-chain length, unsaturation, and positional isomerism further complicate the molecular identification of PLs, while quantitative analysis may be influenced by class-dependent ionization behavior and matrix-induced effects (55, 58). PLs classes also differ in charge properties, where PC and SM are neutral zwitterions, whereas PI, PS, PG, PA, and CL are anionic. These charge and polarity differences define PLs’ solubility and chromatographic behavior and directly affect the extraction and analysis process (29, 59, 60). As a result, this complex structural organization of PLs typically requires the use of mixed solvent systems capable of disrupting both polar and nonpolar interactions (29, 59). However, such approaches often promote co-extraction of other polar constituents, increasing sample complexity and contributing to incomplete recovery, particularly for low-abundance classes such as PI (61). These structure-driven limitations should be considered when analyzing phospholipid compositional data and provide the context for the analytical approaches discussed in section 4 (Figure 3).

Phospholipids are naturally present in foods of animal, plant, and microbial origin, and are common in eggs, milk, meats, fish and shellfish, grains, and oilseeds (Table 1). The most common sources of phospholipids in nature are egg yolk (10.3 g/100 g), beef brain (5.4 g/100 g), and animal livers (2.5–2.9 g/100 g) (8). The phospholipid composition in foods of animal, plant, and microbial origin is variable, as a consequence of the different structures and biological functions of these molecules (29, 60).

Egg yolk is one of the richest natural sources of phospholipids, accounting for approximately 10% of its wet weight and nearly 30% of total yolk lipids. Its PLs composition is characterized by a strong predominance of PD, which typically constitutes more than 70% of total PLs, followed by PE at 15–24%. Minor classes, including LPC, LPE, and SM, together account for only a small fraction of total PLs pool. Owing to this compositional profile, egg yolk is widely recognized as an exceptionally rich source of choline-containing PLs (10, 62, 63).

Milk and dairy products have also been of interest as alternative commercial sources of phospholipids, particularly due to their richness in SM and PS (59). In bovine milk, the PLs fraction is dominated by PE and PC, with substantial contributions from SM, PS, and PI. In human milk, PLs represent a small proportion of total lipids (0.26–0.80%), however, SM is typically the most abundant phospholipid (29–45%) often exceeding PC, with a critical role in supporting infant neurodevelopment and gastrointestinal barrier function (49, 64). The second and third most abundant classes are PC and PE. The reported values for milk phospholipid composition can vary significantly, probably due to methodological differences in lipid extraction and analysis as well as biological factors like breeding period, diet, lactation stage, and technological influences such as processing conditions (28, 65, 66).

Natural phospholipids are extracted primarily from plant-derived and obtained from vegetable sources such as oilseeds (soybeans, sunflower, rapeseed), cereals, wheat germ, flaxseeds, oil-rich nuts, and some legumes Most plant tissues contain only minor amounts of PLs; however, seeds and beans represent the principal reservoirs due to their role in lipid storage (20). Crude vegetable oils contain low but significant PLs content (1–3%), which are further refined to obtain the purified lecithin (25). The dominant PLs classes, across plant-derived materials, are PC, PE, PI, and PS, with PC generally representing the major component of de-oiled vegetable lecithin. Minor constituents such as PA and Lysophospholipids, particularly LPC, are also consistently detected but at much lower levels. The primary and dominant commercial source of PLs is soybean oil (1.5–3% PLs), which yields lecithin with the highest total PLs content (65–75% PLs). PLs composition of plant lecithin is strongly influenced by agronomic variables, genetic background of the crop, environmental conditions, and processing parameters such as extraction and refining. These factors collectively account for the substantial variability reported in PLs subclass distribution across different plant sources and commercial lecithin products (13, 22, 67).

Marine sources, including finfish, shellfish, roe, and algae, represent an important dietary source of PLs, particularly distinguished by their enrichment in long-chain omega-3 polyunsaturated fatty acids such as EPA and DHA (27). Compared with terrestrial PLs, marine-derived PLs typically exhibit a higher degree of unsaturation, with omega-3 PUFAs preferentially esterified at the sn-2 position of the glycerol backbone. The total PLs content of marine organisms varies widely among species and tissues, reflecting differences in physiology and lipid storage. PC is the major class of phospholipid present across marine foods (45–80%), while PE, PI, and PS occur at lower but variable proportions. SM is generally present at relatively low levels in marine species compared with terrestrial animals, although elevated concentrations have been reported in certain shellfish. Notable exceptions to the typical class distribution are observed in specialized tissues such as fish roe, where PI can represent a major PLs class, underscoring its functional importance in reproductive and developmental processes (68, 69). In addition to class composition, marine PLs are characterized by a high degree of fatty acyl unsaturation. PC commonly contains substantial proportions of PUFAs, while DHA is frequently enriched in PE and PS, contributing to the distinctive nutritional and functional properties of marine PLs (29, 70).

Owing to their rapid growth, scalability, and controlled cultivation, microbial systems are increasingly explored as sustainable alternative sources of tailored PLs mixtures with distinct functional properties (71). Microbial sources of phospholipids, such as yeasts, fungi, and bacteria, are characterized by membrane compositions that differ markedly from those of higher plants and animals. Their phospholipid profiles are typically enriched in PE, PI, and CL, reflecting their central roles in maintaining membrane integrity, supporting bioenergetic processes, and enabling adaptation to environmental stress (72, 73).

Mitochondria are distinguished by their cardiolipin-rich membrane anatomy, where CL, PC, and PE comprise the core phospholipid framework (74). CL consistently represents a substantial fraction of the mitochondrial phospholipid pool, highlighting its conserved structural and functional significance in energy metabolism and membrane organization. The relative abundance of these PLs varies across tissues and physiological states (75, 76).

The extraction of food-grade PLs is directly influenced by the noted variability in the origin of each source, class distribution, fatty acyl composition, abundance, and matrix association, as well as the multi-step analysis processes involving concentration, solvent extraction, distillation, and fractionation steps which often result in complex workflows and low yields (41, 51). Plant-derived PLs are typically extracted from oil-rich plants by degumming or solvent fractionation approaches, whereas egg and dairy PLs, which are tightly associated with proteins and membranes, require more complex or multi-step extraction procedures (37, 38, 43, 60, 77). Marine-derived phospholipids are typically high in omega-3 PUFAs, rendering them highly susceptible to oxidative degradation during processing and analysis (29, 70). In contrast, PLs from cereal- and microbe-derived matrices are embedded in polar, structurally complex matrices, where solvent polarity and enzymatic accessibility strongly affect recovery (45, 60, 78). As a conclusion, these source-dependent structural and matrix-specific properties control PLs stability, extraction efficiency, and class selectivity, highlighting the necessity of customized analytical techniques for the PLs sources (29, 35). The implications of these differences for extraction strategy selection and downstream lipidomic analysis are discussed in Section 4.

Recent advances in extraction technologies include both solvent-based and green methods, such as supercritical fluid extraction, ultrasound- or microwave-assisted extraction, and enzyme-assisted extraction (28). The continued development of improved extraction methods is still needed to increase the yield, purity, and eco-friendliness of phospholipid isolation (37, 38, 79). Variations in the lipid classes extracted and reported for food samples can still be observed, and no standard protocol is used for phospholipid extraction and quantitation in foods (37, 38, 54, 79).

After obtaining the phospholipids from the dietary samples, the next significant step is to separate, identify, and quantify them. In food matrices, these steps are quite difficult due to the amphiphilic nature, structural variety, and the requirement to discover small components of phospholipids within complex lipid mixtures. To overcome these problems, a variety of analytical techniques have been developed and are frequently used in tandem, depending on the target PL classes, sample type, and desired sensitivity. Over the years, methodologies have evolved from simple colorimetric tests and thin-layer chromatography (TLC) to more sophisticated techniques like UHPLC, HPLC–MS, and multidimensional lipidomics.

In this section, we tried to summarize the key analytical techniques employed in the separation, identification, quantification, and structural characterization of phospholipids, highlighting the recent advancements that have significantly enhanced sensitivity, resolution, and lipidome coverage. Table 3 presents a comparative summary of different analytical methodologies.

For PLs research and quality control in foods, an effective, and widely applicable analytical framework is: (i) extraction adapted to the matrix of interest; (ii) SPE to remove neutral lipids and pre-enrich PLs; (iii) UHPLC-CAD/ELSD for routine class-level quantification; (iv) LC–MS(/MS) for species-level resolution and tracking of oxidation/Lysophospholipid formation; (v) ³¹P NMR for orthogonal class quantification and method validation; and (vi) GC-FAME for acyl fingerprinting.

This integrated, multi-tiered approach offers analytical depth to be adapted/adjusted according to matrix complexity and study objectives, balancing throughput, specificity, and reproducibility across various food matrices. It also makes the effects of upstream solvent choice on the overall class distribution explicit, facilitating direct comparisons.

Classical solvent-based extraction methods (e.g., Folch and Bligh and Dyer) are still the most frequently used procedures for PLs extraction due to their simplicity and good recoveries. However, the use of toxic organic solvents is becoming an issue in the food and environmental sectors in the search for new “greener” methodologies that minimize the use of hazardous solvents and energy. The recently developed green technologies (ultrasound-, enzyme-, and supercritical fluid–assisted extractions) are characterized by increased selectivity, reduction in solvent use, and higher compatibility with the possible thermal degradation or oxidation of heat- and oxidation-sensitive PLs. These techniques will be further optimized, leading to improved extraction yields with minimal environmental impact and in favor of sustainable production of high-value food, nutraceutical, and pharmaceutical products. The implementation of modern and high-resolution analytical platforms such as UHPLC-CAD, LC–MS/MS, MALDI-TOF, and ³¹P NMR has recently made it possible to precisely profile lipids in a high-throughput mode. However, the lack of reproducibility among these analyses due to the variability in extraction and quantification procedures is a barrier that will have to be overcome in the future. This will require further standardization, automation, and real-time monitoring of the extraction process, which will facilitate comparison between studies. In all cases, the best extraction strategy will depend on the main goal to be achieved (high extraction yield, keeping the native structure, using the lowest amount of energy, or applying a sustainable strategy). Overall, the present paper emphasizes that the selection of PLs extraction and analytical strategies must remain matrix-aware and objective-driven, balancing recovery, molecular integrity, energy efficiency, and sustainability rather than relying on a single universal protocol.